Reflexion of X Rays by Crystals. 885 



over which the crystal is able to reflect. When we com- 

 pare the quantity obtained in, this way with the ionization 

 by the direct primary current the width of the slit dis- 

 appears. Obviously the height of the slit is of no conse- 

 quence to a first order of approximation. In fact, the ratio 

 is only affected now by the times during which the current is 

 allowed to run in each part of the experiment, and the mag- 

 nitude of the steps through which the crystal is turned. As 

 these quantities may be kept constant in different experiments, 

 or altered in known proportions, the ratio is now a true com- 

 parative measure of the intensity of the reflexion of an X ray 

 of given quality by a given crystal face in a given order *. 



In some measurements I have found it convenient to 

 observe the ionization current for a given time at each step 

 in the movement of the crystal, and subsequently to add 

 together all the currents observed. This is of course equiva- 

 lent to plotting the curve and integrating it. At other 

 times I have left the rays in action and advanced the crystal 

 one step at each beat of the clock : the movement of the 

 orystal is begun at an angle at which there is no reflexion, 

 it is taken through the reflexion angle, during which time 

 the leaf of the electroscope moves rapidly, and the rays are 

 turned off when the crystal has ceased to reflect. By this 

 method the integration above mentioned is effected auto- 

 matically. If it were necessary it would be easy to move 

 the crystal at a uniform rate by clockwork, but in practice 

 the simpler methods are quite sufficient for the purpose. 

 The method of steady revolution has been used by de Broglie t 

 in obtaining photographs of the spectra. 



* W7 Lawrence Bragg suggests that this principle is capable of still 

 f urther development. To turn the crystal round at a definite rate is equi- 

 valent to submitting it to radiation from all directions for a given time. 

 Let the intensity of primary radiation within a range 8<p be R8<£, and let 

 dd be the angle which the slit subtends at the point source. (It makes 

 no difference to the argument that the source is not a point, but must 

 be treated as an assemblage of points.) "When the crystal turns round 

 at the rate w, there will be for a time d9/u a ray making any desired 

 angle with the crystal. If the crystal stood still and were irradiated 

 from all directions in the one plane (to consider one plane only does not 

 affect the generality) for a given time, then there would be for that time 

 a ray making any desired angle with the crystal. The experiment there- 

 fore is equivalent to submitting the crystal for a time 8Q/w to radiation E, 

 coming from all directions. If k is the coefficient of reflexion, the in- 

 tensity of the reflected beam may be put equal to k . E, . SQ/u>. On the 

 other hand, the measure of the intensity of the primary beam is R80, if 

 it be allowed to enter the ionization chamber for one second. Hence the 

 ratio of the two observed quantities is k/u> ; and we find a value for k in 

 absolute measure. 



f C. R. Nov. 17, 1913. 



Phil. Mag. S. (5. Vol. 27. No. 161. May 1914. 3 N 



